scLDM: Single-Cell Latent Diffusion Model

scLDM: Single-Cell Latent Diffusion Model#

Overview#

scLDM is a deep learning framework for modeling single-cell gene expression using variational autoencoders (VAEs) and latent diffusion models. The package provides state-of-the-art architectures for learning compressed representations of single-cell transcriptomics data.

Key Components#

Core VAE Architectures#

  • TransformerVAE: Transformer-based VAE architecture for modeling gene expression patterns

Training Modules#

  • VAE: PyTorch Lightning module for training VAE models

  • LatentDiffusion: Latent diffusion model for generative modeling in latent space

Data Handling#

  • DataModule: PyTorch Lightning DataModule for loading and preprocessing single-cell datasets

scldm#

Tests Documentation

single-cell latent diffusion model

Getting started#

Please refer to the documentation, in particular, the API documentation.

Installation#

You need to have Python 3.11 or newer installed on your system. If you don’t have Python installed, we recommend installing uv.

There are several alternative options to install scldm:

  1. Install the latest development version:

pip install git+https://github.com/czi-ai/scldm.git@main

Release notes#

See the changelog.

Contact#

For questions and help requests, you can reach out in the [scverse discourse][]. If you found a bug, please use the issue tracker.

Citation#

t.b.a

Contents